includes/MagicWord.php

   1 <?php
   2 /**
   3  * File for magic words
   4  * @addtogroup Parser
   5  */
   6
   7 /**
   8  * This class encapsulates "magic words" such as #redirect, __NOTOC__, etc.
   9  * Usage:
  10  *     if (MagicWord::get( 'redirect' )->match( $text ) )
  11  *
  12  * Possible future improvements:
  13  *   * Simultaneous searching for a number of magic words
  14  *   * MagicWord::$mObjects in shared memory
  15  *
  16  * Please avoid reading the data out of one of these objects and then writing
  17  * special case code. If possible, add another match()-like function here.
  18  *
  19  * To add magic words in an extension, use the LanguageGetMagic hook. For
  20  * magic words which are also Parser variables, add a MagicWordwgVariableIDs
  21  * hook. Use string keys.
  22  *
  23  */
  24 class MagicWord {
  25         /**#@+
  26          * @private
  27          */
  28         var $mId, $mSynonyms, $mCaseSensitive, $mRegex;
  29         var $mRegexStart, $mBaseRegex, $mVariableRegex;
  30         var $mModified, $mFound;
  31
  32         static public $mVariableIDsInitialised = false;
  33         static public $mVariableIDs = array(
  34                 'currentmonth',
  35                 'currentmonthname',
  36                 'currentmonthnamegen',
  37                 'currentmonthabbrev',
  38                 'currentday',
  39                 'currentday2',
  40                 'currentdayname',
  41                 'currentyear',
  42                 'currenttime',
  43                 'currenthour',
  44                 'localmonth',
  45                 'localmonthname',
  46                 'localmonthnamegen',
  47                 'localmonthabbrev',
  48                 'localday',
  49                 'localday2',
  50                 'localdayname',
  51                 'localyear',
  52                 'localtime',
  53                 'localhour',
  54                 'numberofarticles',
  55                 'numberoffiles',
  56                 'numberofedits',
  57                 'sitename',
  58                 'server',
  59                 'servername',
  60                 'scriptpath',
  61                 'pagename',
  62                 'pagenamee',
  63                 'fullpagename',
  64                 'fullpagenamee',
  65                 'namespace',
  66                 'namespacee',
  67                 'currentweek',
  68                 'currentdow',
  69                 'localweek',
  70                 'localdow',
  71                 'revisionid',
  72                 'revisionday',
  73                 'revisionday2',
  74                 'revisionmonth',
  75                 'revisionyear',
  76                 'revisiontimestamp',
  77                 'subpagename',
  78                 'subpagenamee',
  79                 'displaytitle',
  80                 'talkspace',
  81                 'talkspacee',
  82                 'subjectspace',
  83                 'subjectspacee',
  84                 'talkpagename',
  85                 'talkpagenamee',
  86                 'subjectpagename',
  87                 'subjectpagenamee',
  88                 'numberofusers',
  89                 'rawsuffix',
  90                 'newsectionlink',
  91                 'numberofpages',
  92                 'currentversion',
  93                 'basepagename',
  94                 'basepagenamee',
  95                 'urlencode',
  96                 'currenttimestamp',
  97                 'localtimestamp',
  98                 'directionmark',
  99                 'language',
 100                 'contentlanguage',
 101                 'pagesinnamespace',
 102                 'numberofadmins',
 103                 'defaultsort',
 104         );
 105
 106         static public $mObjects = array();
 107
 108         /**#@-*/
 109
 110         function __construct($id = 0, $syn = '', $cs = false) {
 111                 $this->mId = $id;
 112                 $this->mSynonyms = (array)$syn;
 113                 $this->mCaseSensitive = $cs;
 114                 $this->mRegex = '';
 115                 $this->mRegexStart = '';
 116                 $this->mVariableRegex = '';
 117                 $this->mVariableStartToEndRegex = '';
 118                 $this->mModified = false;
 119         }
 120
 121         /**
 122          * Factory: creates an object representing an ID
 123          * @static
 124          */
 125         static function &get( $id ) {
 126                 if (!array_key_exists( $id, self::$mObjects ) ) {
 127                         $mw = new MagicWord();
 128                         $mw->load( $id );
 129                         self::$mObjects[$id] = $mw;
 130                 }
 131                 return self::$mObjects[$id];
 132         }
 133
 134         /**
 135          * Get an array of parser variable IDs
 136          */
 137         static function getVariableIDs() {
 138                 if ( !self::$mVariableIDsInitialised ) {
 139                         # Deprecated constant definition hook, available for extensions that need it
 140                         $magicWords = array();
 141                         wfRunHooks( 'MagicWordMagicWords', array( &$magicWords ) );
 142                         foreach ( $magicWords as $word ) {
 143                                 define( $word, $word );
 144                         }
 145
 146                         # Get variable IDs
 147                         wfRunHooks( 'MagicWordwgVariableIDs', array( &self::$mVariableIDs ) );
 148                         self::$mVariableIDsInitialised = true;
 149                 }
 150                 return self::$mVariableIDs;
 151         }
 152
 153         # Initialises this object with an ID
 154         function load( $id ) {
 155                 global $wgContLang;
 156                 $this->mId = $id;
 157                 $wgContLang->getMagic( $this );
 158                 if ( !$this->mSynonyms ) {
 159                         $this->mSynonyms = array( 'dkjsagfjsgashfajsh' );
 160                         #throw new MWException( "Error: invalid magic word '$id'" );
 161                         wfDebugLog( 'exception', "Error: invalid magic word '$id'\n" );
 162                 }
 163         }
 164
 165         /**
 166          * Preliminary initialisation
 167          * @private
 168          */
 169         function initRegex() {
 170                 #$variableClass = Title::legalChars();
 171                 # This was used for matching "$1" variables, but different uses of the feature will have
 172                 # different restrictions, which should be checked *after* the MagicWord has been matched,
 173                 # not here. - IMSoP
 174
 175                 $escSyn = array();
 176                 foreach ( $this->mSynonyms as $synonym )
 177                         // In case a magic word contains /, like that's going to happen;)
 178                         $escSyn[] = preg_quote( $synonym, '/' );
 179                 $this->mBaseRegex = implode( '|', $escSyn );
 180
 181                 $case = $this->mCaseSensitive ? '' : 'iu';
 182                 $this->mRegex = "/{$this->mBaseRegex}/{$case}";
 183                 $this->mRegexStart = "/^(?:{$this->mBaseRegex})/{$case}";
 184                 $this->mVariableRegex = str_replace( "\\$1", "(.*?)", $this->mRegex );
 185                 $this->mVariableStartToEndRegex = str_replace( "\\$1", "(.*?)",
 186                         "/^(?:{$this->mBaseRegex})$/{$case}" );
 187         }
 188
 189         /**
 190          * Gets a regex representing matching the word
 191          */
 192         function getRegex() {
 193                 if ($this->mRegex == '' ) {
 194                         $this->initRegex();
 195                 }
 196                 return $this->mRegex;
 197         }
 198
 199         /**
 200          * Gets the regexp case modifier to use, i.e. i or nothing, to be used if
 201          * one is using MagicWord::getBaseRegex(), otherwise it'll be included in
 202          * the complete expression
 203          */
 204         function getRegexCase() {
 205                 if ( $this->mRegex === '' )
 206                         $this->initRegex();
 207
 208                 return $this->mCaseSensitive ? '' : 'iu';
 209         }
 210
 211         /**
 212          * Gets a regex matching the word, if it is at the string start
 213          */
 214         function getRegexStart() {
 215                 if ($this->mRegex == '' ) {
 216                         $this->initRegex();
 217                 }
 218                 return $this->mRegexStart;
 219         }
 220
 221         /**
 222          * regex without the slashes and what not
 223          */
 224         function getBaseRegex() {
 225                 if ($this->mRegex == '') {
 226                         $this->initRegex();
 227                 }
 228                 return $this->mBaseRegex;
 229         }
 230
 231         /**
 232          * Returns true if the text contains the word
 233          * @return bool
 234          */
 235         function match( $text ) {
 236                 return preg_match( $this->getRegex(), $text );
 237         }
 238
 239         /**
 240          * Returns true if the text starts with the word
 241          * @return bool
 242          */
 243         function matchStart( $text ) {
 244                 return preg_match( $this->getRegexStart(), $text );
 245         }
 246
 247         /**
 248          * Returns NULL if there's no match, the value of $1 otherwise
 249          * The return code is the matched string, if there's no variable
 250          * part in the regex and the matched variable part ($1) if there
 251          * is one.
 252          */
 253         function matchVariableStartToEnd( $text ) {
 254                 $matches = array();
 255                 $matchcount = preg_match( $this->getVariableStartToEndRegex(), $text, $matches );
 256                 if ( $matchcount == 0 ) {
 257                         return NULL;
 258                 } else {
 259                         # multiple matched parts (variable match); some will be empty because of
 260                         # synonyms. The variable will be the second non-empty one so remove any
 261                         # blank elements and re-sort the indices.
 262                         # See also bug 6526
 263
 264                         $matches = array_values(array_filter($matches));
 265
 266                         if ( count($matches) == 1 ) { return $matches[0]; }
 267                         else { return $matches[1]; }
 268                 }
 269         }
 270
 271
 272         /**
 273          * Returns true if the text matches the word, and alters the
 274          * input string, removing all instances of the word
 275          */
 276         function matchAndRemove( &$text ) {
 277                 $this->mFound = false;
 278                 $text = preg_replace_callback( $this->getRegex(), array( &$this, 'pregRemoveAndRecord' ), $text );
 279                 return $this->mFound;
 280         }
 281
 282         function matchStartAndRemove( &$text ) {
 283                 $this->mFound = false;
 284                 $text = preg_replace_callback( $this->getRegexStart(), array( &$this, 'pregRemoveAndRecord' ), $text );
 285                 return $this->mFound;
 286         }
 287
 288         /**
 289          * Used in matchAndRemove()
 290          * @private
 291          **/
 292         function pregRemoveAndRecord( ) {
 293                 $this->mFound = true;
 294                 return '';
 295         }
 296
 297         /**
 298          * Replaces the word with something else
 299          */
 300         function replace( $replacement, $subject, $limit=-1 ) {
 301                 $res = preg_replace( $this->getRegex(), StringUtils::escapeRegexReplacement( $replacement ), $subject, $limit );
 302                 $this->mModified = !($res === $subject);
 303                 return $res;
 304         }
 305
 306         /**
 307          * Variable handling: {{SUBST:xxx}} style words
 308          * Calls back a function to determine what to replace xxx with
 309          * Input word must contain $1
 310          */
 311         function substituteCallback( $text, $callback ) {
 312                 $res = preg_replace_callback( $this->getVariableRegex(), $callback, $text );
 313                 $this->mModified = !($res === $text);
 314                 return $res;
 315         }
 316
 317         /**
 318          * Matches the word, where $1 is a wildcard
 319          */
 320         function getVariableRegex()     {
 321                 if ( $this->mVariableRegex == '' ) {
 322                         $this->initRegex();
 323                 }
 324                 return $this->mVariableRegex;
 325         }
 326
 327         /**
 328          * Matches the entire string, where $1 is a wildcard
 329          */
 330         function getVariableStartToEndRegex() {
 331                 if ( $this->mVariableStartToEndRegex == '' ) {
 332                         $this->initRegex();
 333                 }
 334                 return $this->mVariableStartToEndRegex;
 335         }
 336
 337         /**
 338          * Accesses the synonym list directly
 339          */
 340         function getSynonym( $i ) {
 341                 return $this->mSynonyms[$i];
 342         }
 343
 344         function getSynonyms() {
 345                 return $this->mSynonyms;
 346         }
 347
 348         /**
 349          * Returns true if the last call to replace() or substituteCallback()
 350          * returned a modified text, otherwise false.
 351          */
 352         function getWasModified(){
 353                 return $this->mModified;
 354         }
 355
 356         /**
 357          * $magicarr is an associative array of (magic word ID => replacement)
 358          * This method uses the php feature to do several replacements at the same time,
 359          * thereby gaining some efficiency. The result is placed in the out variable
 360          * $result. The return value is true if something was replaced.
 361          * @static
 362          **/
 363         function replaceMultiple( $magicarr, $subject, &$result ){
 364                 $search = array();
 365                 $replace = array();
 366                 foreach( $magicarr as $id => $replacement ){
 367                         $mw = MagicWord::get( $id );
 368                         $search[] = $mw->getRegex();
 369                         $replace[] = $replacement;
 370                 }
 371
 372                 $result = preg_replace( $search, $replace, $subject );
 373                 return !($result === $subject);
 374         }
 375
 376         /**
 377          * Adds all the synonyms of this MagicWord to an array, to allow quick
 378          * lookup in a list of magic words
 379          */
 380         function addToArray( &$array, $value ) {
 381                 global $wgContLang;
 382                 foreach ( $this->mSynonyms as $syn ) {
 383                         $array[$wgContLang->lc($syn)] = $value;
 384                 }
 385         }
 386
 387         function isCaseSensitive() {
 388                 return $this->mCaseSensitive;
 389         }
 390 }
 391
 392 ?>